| Commit message (Collapse) | Author | Age | Files | Lines |
|
|
|
|
| |
* use generics in random
* fix
|
| |
|
|
|
|
|
| |
* use funcs and inline in strutils
* use funcs
|
|
|
|
|
| |
* fix parseChar
* do not introduce new double backticks
|
|
|
|
|
|
|
|
|
| |
The default HTML template for nimdoc currently initializes the dark mode switch and pragma dots when the onload event is fired. But since the onload event does not fire until all external resources (images, stylesheets, fonts, etc.) have been loaded, the light theme is shown for a brief moment before the document is fully loaded, and it switches to the dark theme. This is quite jarring, especially on slower internet connections. So let's instead initialize these things on the DOMContentLoaded event, which is fired right after the document has been parsed and the initial DOM structure is ready. This means that we now display the dark mode immediately, without having to wait for external resources to load first.
For reference, see:
- https://developer.mozilla.org/en-US/docs/Web/API/Window/load_event
- https://developer.mozilla.org/en-US/docs/Web/API/Window/DOMContentLoaded_event
I have updated the snapshot tests in `nimdoc/`, and done some manual testing of both the theme switch and the pragma dots, to confirm that this does not break anything. Please let me know if I've missed anything.
|
|
|
|
|
|
|
|
|
| |
* [docs minor]space for code-block
* correct more errors
* to runnableExamples
* add newline
|
|
|
|
| |
* better docs and tests
* a bit better only clean trailing whitespace
|
| |
|
|
|
|
|
|
|
| |
consumption (#16250)
* ORC: make the adaptive strategy the default in order to fight memory consumption
* added missing test case
|
| |
|
| |
|
| |
|
| |
|
| |
|
|
|
|
|
| |
* js module also uses runnableExamples
* cleanup docs and tests
|
|
|
|
| |
* changelog
* add testcase, fixes #14332
|
| |
|
| |
|
|
|
|
|
|
|
|
|
| |
* change break->continue and rewrite one-yield-style
* use case statement for clarity
* Tiny nit
Co-authored-by: Clyybber <darkmine956@gmail.com>
|
|
|
|
|
|
|
|
|
| |
`addr(mystring[ind])` (index + index assignment) (#15987)
* fix #15939, fix #15464 VM now supports `addr(mystring[ind])` (index + index assignment), var char return etc
* cleanups
* cstring tests
* add test for bug #15464
* improve test coverage
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
* js module also uses runnableExamples
* better setops docs
* Update lib/system/setops.nim
* better
* Update lib/system/setops.nim
* Update lib/system/setops.nim
* better
* better
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
* Fixes #16219, `hasArgOfName` ignoring argument sets.
* Fix test and simplify ident traversal.
* Moved test into a block and removed some boilerplate.
* Fix some argument formatting.
* use ..<
* Change the preceding line too
Co-authored-by: Clyybber <darkmine956@gmail.com>
|
|
|
|
|
|
|
|
|
|
|
|
| |
* try to fix #16206
* merge two parsecfg tests into one
* Revert "fix"
This reverts commit 668bdec2c499cf9967abfb7aad24975a04b092eb.
* Revert "try to fix #16206"
This reverts commit c399cc2153190299c9cbb0ad83f6ce85a7bbbb89.
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
With this commit, we no longer see an error if we pass
`--styleCheck:error` when compiling a file that contains `import sugar`
or `import std/with`.
The problem was that those modules (and only those modules) import
`std/private/underscored_calls`, which contained a styleCheck issue:
its spelling of `nnkArgList` didn't match the `nnkArglist` spelling in
`macros.nim`.
This commit fixes the issue by renaming `nnkArgList` to `nnkArglist`
repo-wide. The other way around would be a breaking change for code that
used `nnkArglist` and `--styleCheck:error`.
Fixes: #16174
|
| |
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
* [docs minor]better comparisons docs
* switch from PRNG to random module
* apply suggestions
* not compile
* Revert "switch from PRNG to random module"
This reverts commit 83b4d8946d7d677edac43b7675c41e230ba4e382.
* small style changes and clarify clamp
|
| |
|
|
|
|
|
| |
* fix #16164, render doc comments
* add a test
|
| |
|
| |
|
| |
|
| |
|
|
|
|
|
|
|
| |
* Don't add content-length header on GET request when there is no body.
* Add space between the not and the (.
* Take Dom's suggestion.
|
|
|
|
|
| |
* ORC: API extensions
* ORC: exploit a common special case
|
| |
|
| |
|
| |
|
|
|
|
|
|
| |
* better addint
* Update lib/system/strmantle.nim
Co-authored-by: Juan Carlos <juancarlospaco@gmail.com>
|
|
|
|
|
|
|
| |
* fix broken => for typed procs with pragmas
* add more sugar tests
* add ending quote
|
|
|
| |
This reverts commit bb4b27a2ca414f06fbb9d14ff76fa02a088ac141.
|
|
|
|
|
|
|
|
|
|
|
| |
* fix #16185
* fix test
* fix comment
* fix comment
* better approach
|
|
|
|
|
|
|
|
|
|
|
|
| |
* toXXAscii use xor op, saving 30%~50% time
* Update lib/pure/strutils.nim
Co-authored-by: hlaaftana <10591326+hlaaftana@users.noreply.github.com>
* Update lib/pure/strutils.nim
Co-authored-by: Andreas Rumpf <rumpf_a@web.de>
Co-authored-by: hlaaftana <10591326+hlaaftana@users.noreply.github.com>
|
| |
|
| |
|
| |
|
| |
|
| |
|
| |
|
|
|
|
|
|
|
| |
--targets (#16163)
* fix #14242 `testament r tests/js/foo` now works; testament now honors --targets
* fix shouldfail
|